Technology Treading Lightly

The Power of Incentives on Data Center Efficiency

Mark Aggar — Fri, 30 Mar 2012 22:26:23 GMT

I recently had the honor of having a blog post published on Corporate EcoForum's EcoInnovator blog that focused on the role of incentives and data center efficiency. Take a look:

http://corporateecoforum.com/ecoinnovator/?p=7035

As incredible as it sounds, and as I wrote in The IT Energy Efficiency Imperative, a typical data center uses less than 5% of the energy it consumes for actual computing –the rest is lost due to various overheads and inefficiencies. If that wasn’t bad enough, data centers also generate huge piles of e-waste each year as computer servers are replaced by more powerful and, ironically, more energy efficient models.

While this may sound terrible, first keep in mind that it usually takes less energy and natural resources to communicate and consume goods digitally rather than physically, so more investment in IT is often net positive for the environment.

However, underlying all those boasts about new, “energy efficient” data centers, there is usually a dirty little secret: the servers are often woefully under-utilized. Most work at less than 10 percent of their capacity, but suck up about half of the energy that they would when running at full capacity. So all over the world, millions of servers operate like mostly empty delivery trucks, consuming lots of resources, but delivering very little.

True, most IT departments have tried to tackle poor server utilization by consolidating applications onto fewer servers through the use of virtualization technology. Unfortunately, many consolidation projects have stalled due to financial constraints, organizational politics and staffing shortages. As such, server utilization across the enterprise remains very low – often less than 10 percent. Even worse, as each generation of server hardware has become more powerful, utilization has tended to decrease even further as applications and IT operational practices have been unable to take advantage of it.

Continued...

Windows Azure Autoscaling–available now!

Mark Aggar — Fri, 09 Dec 2011 00:16:00 GMT

In my white paper on IT energy efficiency, I wrote about the need for software developers to design their applications to dynamically scale, so we aren’t simply trading idle servers for idle VMs when applications move to public or private clouds. Unlike physical server infrastructure which is often statically assigned to specific applications for years, private and public clouds typically provide incentives for developers to scale up and down dynamically because the billing increment can be quite small – often by the hour. So you can control costs and reduce the environmental impact of your apps at the same time – if you design them appropriately!

Up until now, scaling instances on Windows Azure was a manual process or required you to develop your own functionality to do this automatically.

In response to this opportunity for improved resource efficiency and utilization, the Microsoft Patterns & Practices team has worked with Microsoft Consulting Services and an advisory board of Windows Azure customers to develop the Windows Azure AutoScaling Block, which RTW’ed last week as part of the new Enterprise Library Integration Pack for Windows Azure.

WASABi, as it’s affectionately known, helps developers to automatically scale both web and worker roles in Windows Azure by dynamically provisioning/decommissioning roles or throttling. These scaling actions are based on timetables or on metrics collected from the application and/or Windows Azure Diagnostics.

WASABi addresses the following scenarios:

1. Autoscaling both web and worker roles in Windows Azure by dynamically changing instance counts or performing application throttling.

2. Autoscaling Windows Azure roles based on timetables.

3. Autoscaling Windows Azure roles based on metrics collected from the application and/or Windows Azure but constrained by upper and lower bounds on the instance count per role.

4. Preventing fast oscillations in the number of role instances through the use of a “stabilizer”. The stabilizer can also help to optimize costs by limiting scaling up operations to the beginning of the hour and scaling down operations to the end of the hour.

5. Monitoring and logging autoscaling activity.

6. Sending notifications to preview any scaling operations before they take place.

7. Encrypting the rules and other configuration in Windows Azure blob storage or in local file storage.

8. Managing the autoscaler configuration by using Windows PowerShell.

WASABi scales apps through rules that a developer and/or an IT Pro creates and updates over time. The excellent blog post by Julian Dominguez notes that there are 2 types of rules in WASABi:

Constraint Rules- Defines lower and upper boundaries for the number of instances that should be present at all times. You can define boundaries that apply at all times, and you can also define boundaries based on recurrence patterns, to proactively scale your application based on expected load.
Reactive Rules- Allows you to scale your application based on conditions, such as when a certain performance counter goes beyond a certain level (e.g. average CPU usage for all instances of a role is greater than 80%), when a Windows Azure Queue has too many messages, or if you have your own metric you are interested in, you can extend and support it.

The recommended way to obtain the Enterprise Library Integration Pack for Windows Azure is as NuGet packages. Alternatively, you can download self-extracting zip files with binaries, sources (including tests) and the reference implementation from MSDN. The reference implementation also includes an Autoscaling management site, which lets you do authoring of the rules, and monitoring of the autoscaled application, so you can see how Tailspin (the reference implementation application) responds to load. The configuration tool is available as a Visual Studio extension package (VSIX) from the Visual Studio Gallery.

The buzz and uptake on WASABi has been very positive thus far and which leads me to believe that WASABi will be a key ingredient for any Windows Azure app that expects (or hopes) to scale up significantly. Next up – Pickled Ginger?

P.S. If you were wondering, the root pictured above is the real Wasabi root, which you can read more about here.

The IT Energy Efficiency Imperative

Mark Aggar — Wed, 08 Jun 2011 17:07:20 GMT

Once again, I've had a long hiatus from blogging. But I've got an excuse (this time!). I've been heads down on finalizing a new white paper on IT energy efficiency. I also just returned from Corporate EcoForum's annual meeting where a number of the attendees confirmed the key assertion in the paper - that in most organizations, server utilization is embarassingly low and falling, and (as the chart above shows) improving it provides the biggest opportunity for increasing IT energy efficiency. Move over PUE! (we still love you though).

As I posted on Microsoft's sustainability blog, the paper explores the critical importance and substantial benefits of embracing IT energy efficiency in a world where almost every facet of business and society is dependent on IT-based services. The exponential demand for these IT services is beginning to strain the capacity and cost effectiveness of many organizations’ IT infrastructures. It shows why adopting strong IT energy efficiency principles and practices – from the top of the organization down to the IT pro - can significantly mitigate these challenges and improve the cost effectiveness, capacity and flexibility of tomorrow’s IT infrastructures.

In this paper, we:

Highlight the financial, productivity and environmental benefits of embracing IT energy efficiency.
Illustrate why increasing IT resource utilization from today’s low levels offers the most significant energy efficiency gains and why current efforts to improve utilization are failing.
Provide motivation and actionable guidance for IT and business decision makers to improve IT energy efficiency and resource utilization through technologies such as cloud computing.

If you are looking to reduce costs and increase the ROI from IT (particularly within data centers), or simply contribute to reducing the environmental impact of IT, please download this onscreen-readable whitepaper and let me know what you think.

IE9 Energy efficiency

Mark Aggar — Tue, 29 Mar 2011 23:50:24 GMT

The IE team recently published the results from a set of tests comparing the energy efficiency of the latest competing browsers. Spoiler-alert - IE9 was the most efficient. :-)

http://blogs.msdn.com/b/ie/archive/2011/03/28/browser-power-consumption-leading-the-industry-with-internet-explorer-9.aspx

Big time savings with System Center Configuration Manager 2007 R3

Mark Aggar — Wed, 22 Sep 2010 00:50:00 GMT

The System Center team blog featured a guest post by one of their MVPs yesterday on his experiences with the centralized power management features in SCCM 07 R3.

While not the only feature in the release, the centralized power management functionality certainly seems to be getting the most amount of airtime. Which is not really surprising given that the savings gained from implementing power management would seem to easily pay for the product many times over. As Panu Saukko reports, his customer was spending about $2500/weekday for electricity without power management. In early September they enforced power management on 5000 computers using SCCM 2007 R3, and decreased the power bill by to about $300 per day! And this was with fairly lenient settings - 3 hours of idle before automatic sleep during the day, which means that most PCs are on all day during the work day. The opportunity for more power savings is still significant as they crank down the idle-to-sleep timeouts.

Read the post on the System Center blog here.

Study: New ENERGY STAR rated server consumes 54% less energy than older server

Mark Aggar — Mon, 20 Sep 2010 19:05:52 GMT

Microsoft has been working with the US EPA and HP for the last 6 months or so on a study to compare a new ENERGY STAR-qualified server (2009 model) with a similar, older non-energy star rated servers (2006 model) The results show quite convincingly that replacing an older server with a new ENERGY STAR-qualified model will save energy and deliver more processing power in the bargain. Also, this is a nice demonstration of the work we did with Intel and HP in Windows Server 2008 R2 to improve power management vs. our previous server OSes.

I analyzed some of the raw data from the tests and discovered that in addition to saving over 1/2 the power on idle (54%) as mentioned in the study, the newer server provides almost 300% more transactions at any given CPU load than the 2006 model year server, resulting in about a 300-400% improvement across the board on performance-to-power ratios.

This is really great news. Of course, it also implies that in order to really take advantage of these new servers, you must make sure you are really driving up utilization or that extra performance is going to be wasted. Simply consolidating even more idle VMs to these new servers really isn’t going to cut it. Your ops management infrastructure needs to be aware of what is going on inside those VMs so it can shut some of them down if appropriate. This is going to require applications to take a more active role in resource utilization in the future.

Download the study here (2.67MB)

The Efficiency Imperative

Mark Aggar — Wed, 15 Sep 2010 23:07:00 GMT

On Monday, I was proud to speak on behalf of Microsoft at Intel Developer Forum 2010 on the critical need for developers to do their part to help improve IT efficiency. IT efficiency is a component of Greener IT, which is the term Microsoft is using for recognizing that Green IT is a journey rather than a destination.

As I laid out in my presentation, IT efficiency has (or should) become the next critical focus for IT. Efficiency shares characteristics similar to security, in that everyone involved, including Business Decision Makers, developers, IT Pros, and even users, together with hardware and software vendors, need to do their part to improve IT efficiency.

Readers of my blog know that applications have been the missing link that prevents PCs and servers from being highly efficient. Now is the time to fill that void. With Windows 7 and Windows Server 2008 R2, IT Pros, developers and enthusiasts are able to use in-box diagnostic tools to determine why power management is not working as expected. Microsoft uses these tools every day to identify problems with its own and even third party applications, and we are asking the rest of the Windows ecosystem to do the same to ensure that their applications are “energy smart”.

So where should a developer start? During my talk, I unveiled the “Energy Smart Software Top 10”, which provides recommendations for how developers should prioritize their energy smart software efforts:

For developers of client software, steps 1 through 8 are relevant. For server developers 3-10 are relevant. Additionally, I outlined some of the key investments that will help keep systems idle.

We have published a collection of guidance on www.microsoft.com/energysmart that covers the majority of these recommendations to help developers build energy smart software. More guidance will be added over the coming weeks and months.

There are also links on the site to tools, including an update of the Joulemeter research project by Microsoft Research that we also announced at IDF. The new version of Joulemeter helps democratize computer power measurement by allowing laptop users to use their laptop battery to calibrate the power model to their specific machine.

Developers who have questions on developing energy smart applications can post questions in our MSDN forum.

While there are many factors that are forcing the need to focus on IT efficiency (e.g. increased demand for IT, power constraints, regulations and stakeholder pressure), there are a lot of benefits as well. Users can be more productive due to better power management experiences and TCO can be reduced through smaller energy bills and less idle equipment. And it won’t hurt that our planet’s life support systems will be just a little less stressed as we slow down the rate of extraction and waste disposal (including carbon into the atmosphere) as a result of more efficient IT utilization.

As I closed my presentation, I noted that Intel calculated that the world’s 1 billion PCs collectively consumed more than 320 terawatt hours of electricity in 2007. They estimate that by 2014, advances in computing will allow 2 billion PCs to consume less than half of that amount (151 twH), but provide more than 17 times the computing power.

This is certainly would be an impressive feat for energy-efficient performance, and one that I hope the industry can meet. But, as I pointed out, hardware and operating system advances alone will not get us there. Developers will need to take an active role and ensure their software is energy smart if we are to get close to achieving this goal.

Again, if you write software for a living, you owe it to yourself, your customers and the environment to become educated on energy smart software development. Please visit http://microsoft.com/energysmart today!

For those of you who couldn’t attend my talk, a recorded interview of me talking about the topic is available on MSDN’s channel 9. If and when we get the video from IDF, I’ll post a link to that too.

Windows 7 Power Use Study

Mark Aggar — Wed, 07 Jul 2010 00:39:35 GMT

In the category of “I wrote this a long time ago but forgot to post it”, Mindteck updated their 2009 white paper which explores power use of various PC platforms using Windows 7. In the new paper, they tested both sleep, idle, low use and high use scenarios and built a model to estimate cost savings by using a centralized power management policy.

Possibly one of the most interesting tests they did was the effect of processor chipset drivers on the power consumption of Windows XP and Windows 7. (table here is copy and pasted from the appendix of the doc, and painfully reformatted)

OS	Windows XP			Windows 7			% Improvement Win7
PC Configuration	Idle	Low	High	Idle	Low	High	Idle	Low	High
P4 Updated Drivers	64.2	69.7	89.8	57.3	66.1	79.4	10.75%	5.16%	11.58%
P4 Out-of-box	64.2	68.7	106.2	57.3	66.1	79.4	10.75%	3.78%	25.24%
High-end Updated Drivers	47.2	48.0	67.7	45.2	49.1	66.8	4.14%	2.29%	1.33%
High-end Out-of-box	50.5	54.3	78.0	45.2	49.1	66.8	10.50%	9.54%	14.36%

What is notable is that while W7 beats XP on every test, the “out of box” differences are really significant. If you look closely, Windows 7 out of the box numbers are the same as with the updated drivers – this means that Windows 7 is taking care of the chipset drivers, even on older hardware. The same cannot be said for XP. And even with updated drivers (obtained manually), it’s still worse than Windows 7. How many folks have reinstalled Windows XP and not installed the chipset drivers. Quite a few I would guess.

The Mindteck green computing white paper can be obtained at http://www.mindteck.com/whitepapers/Enabling-Green-Computing.pdf

Getting started with Windows Power Management

Mark Aggar — Sat, 19 Jun 2010 00:24:10 GMT

My team just published three new “getting started” white papers on microsoft.com today. The first paper is an overview of the importance of ensuring that your application, service or driver is “energy smart”. The other two cover some high level best practices for being compatible with system sleep and display power management, and writing efficient code and the associated use of timers.

If you’re just getting started on making your software “energy smart”, I strongly encourage you to take a look. Enjoy!

(Faulty?) Soft power switch caused resume from sleep failures

Mark Aggar — Tue, 08 Jun 2010 01:06:00 GMT

When the motherboard (or ‘mobo’) in my ‘media’ PC stopped working recently, I decided to turn lemons into lemonade and use it as an excuse to build a Home Theatre PC for my living room. While I was deciding on which HTPC case to buy, I installed the new motherboard in my old case to test it out.

My original media PC used to frequently fail to resume from sleep while unattended (e.g. waking up to record a TV show). It would just continually initialize the DVD drive and the hard drive light would flash on and off. I had tried changing just about everything to resolve this problem (power supply, graphics, memory). In the end suspected it was either the mobo or the processor. As a workaround, I configured the machine to bypass S3 sleep and go straight to S4 hibernate, which while it worked, was slow compared to S3 sleep and made for a long wait when firing up the XBOX 360 in Media Center extender mode.

I figured that I had a good chance of resolving this problem for good with the new mobo. For a short while things looked promising, but it wasn’t long before I noticed the same failure as I’d seen with my old mobo!

Given that the processor was also new, all signs pointed to the case. There really wasn’t anything special about it (e.g no media readers, or other onboard goodies except a couple of front USB connections). So on a hunch, I swapped the power and reset connections on the motherboard so that the reset button would power on the machine instead.

My new (temporary) setup never experienced any resume from sleep failures after that (I also ran the system through about 1000 sleep/resume cycles using pwrtst.exe from the Windows Driver Kit to make sure). What was odd is that I’d never experienced any problems when manually resuming from sleep, so I have no idea how the power switch was randomly interfering with unattended wake.

And at the end of the day, I don’t really care now that S3 sleep is working reliably!

Found: Secret Internet Explorer Insomnia feature!

Mark Aggar — Sat, 05 Jun 2010 00:39:00 GMT

In the category of “this problem has been around for years but no-one has noticed it”, I discovered very recently that IE raises an availability request when it downloads a file. This is the desired behavior – you don’t want the PC to automatically sleep during the middle of a download. However, once the download is finished, the availability request isn’t dropped until after you dismiss the ‘download completed’ dialog box. So for large downloads that you start before you go to bed, be sure to check the box that automatically closes the download dialog box so your PC can enjoy a good night’s rest when it’s done downloading. A bug has been filed against IE 9.

Tales of PC Insomnia, Part 2

Mark Aggar — Fri, 26 Mar 2010 00:40:00 GMT

In my last post on this topic, I described a couple of scenarios in which your PC would not automatically sleep. I’ve got good news to report on both.

Streaming Web Video

The first is our old friend Adobe Flash. I’ve been working with the Adobe on this issue for some time now, and with some help from a well known environmental NGO and a large user of Flash, the great news is that it appears that most of the issues are now fixed with the current beta 3 release of Flash 10.1. Before you all rush to go and download it there are a few things you should know.

First the good news (yes there’s a catch). Flash 10.1b3 now appears to close the audio channel when a streaming video ends or is paused. This is important because previous versions of Flash streamed silence to the audio stack, which in turn prevented Windows from automatically putting the PC to sleep. Additionally, when viewing Flash videos full screen, Flash prevents the screen from dimming or blanking, which is great if you are sitting back and watching the screen. It will not prevent dimming if you do not put the viewer in full screen mode. I’m guessing that most folks watch short videos in non-full screen mode, so dimming may still be an issue for some.

The bad news (for now) is that Flash changes the timer resolution down to 1ms as soon as a web page loads it. What this means is that it uses more power than it needs to and in particular will drain your battery faster. (I’ve heard up to 20% faster but have not confirmed this with my own measurements). However, I’ve been assured that this is a temporary situation and that Flash will revert back to it’s old 10.0 behavior in this regard for the final release (it’s there just for the beta I’m told).

So, my advice is that if you have a desktop or a laptop that spends most of its life plugged into the wall, its probably better to install 10.1b3 so automatic sleep has a chance to work more often (keep in mind that this is a beta, and you may run into other non-desirable issues).

If you run on battery a lot and frequently browse the web, you might want to think twice about installing 10.1 until it’s officially released. Unfortunately I have no idea when that will be, but I’m hoping it’s not too far out! Even if you don’t think you use Flash much while on battery, you’d be surprised how often sites cause it to load up just in case. Even MSN.com loads up Flash when you visit, even though it doesn’t appear to use it.

Open files across the network

After discovering that Microsoft Office PowerPoint and Excel seemed to keep files open across the network. which in turn prevented Windows from automatically sleeping, I started a thread with both the Office and Offline folders team. It turns out that there are two issues here, one that is getting fixed in Windows 7 SP1 and another that you can fix yourself.

The first problem was a bug in Offline folders (the files that PowerPoint and Excel were opening were mistakenly seen as ‘sparse’ but Offline folders, which prevented the machine from sleeping). This is great if you are using Offline Folders and deploy SP1 whenever that is ready, but what can you do about it now?

Well it so happens there is a workaround to this and I’ve been running it for several months to make sure there were no ill effects. It involves changing a Windows 7 power policy setting, and while quite effective is not particularly easy to change as it involves reading and writing multiple GUIDs. So I created a batch file which automatically changes the policy on the current active power plan (e.g. if you’re running the balanced plan and then run the batch file, the policy is changed for that power plan).

Before I get to the batch file which I will paste here, a little bit on what the policy does and what impact it has.

The policy is actually a hidden policy called “Allow Sleep with remote opens”. When enabled this allows the PC to automatically sleep even when files are open read-only across the network. Unless PowerPoint or Excel is actively writing a file, the file is open read-only the rest of the time.

The side effect of this policy is that if you computer goes to sleep when you have a PPT or XLS file open and you have made changes to it but not saved, then you will occasionally get a message from the application that the changes that were made have not been saved and that you should save the changes in a different file.

I used to occasionally see these messages before I changed this policy, so this isn’t really a big deal. Simply save the file as a new file and continue.

I’m hoping to have Microsoft IT run this policy against a large number of machines inside Microsoft to see if there are any other side effects. My hope is that if all turns out well that this becomes the default behavior for Windows 8. Until then, you can change the policy yourself and let me know what your experiences are. Keep in mind that the only files I typically have open across the network are opened by Microsoft applications, so I don’t know what it will do with other apps. Your mileage may vary and the standard legal disclaimer applies here.

Before you run the batch file to change the policy, open up a PowerPoint or Excel file across the network and do a ‘powercfg /requests (see my last “Insomnia” post for how to do that if necessary). You should see the open file blocking automatic sleep. Close the file and run the the batch file, and then open up the file again. You should see that the open file is not blocking automatic sleep anymore.

With that said, here’s the batch file. Copy and paste this into notepad, save it as a .cmd file (I call it cfopp.cmd) and then double click on it (make sure you don’t accidentally save it with a .txt extension on the end, its easy to do and of course it won’t run).

Note this is only tested on Windows 7. I have not tried this on Vista (not sure if the policy even exists on Vista). It definitely won’t work on XP. Also, my ‘success’ test is a little fragile, so do a powercfg /qh if you see the failure message and check to see if the “allow sleep with remote opens” policy is set to 1 or 0. If it’s set to 1, you’re all set.

Use the comments link to let me know how this works out for you.

@echo off
REM CFOPP - Change File Open Power Policy
REM Change policy so automatic sleep can occur with remote file opens
REM v1.0

set currentpowerscheme=
set currentpowerschemeguid=
set sleepsubgroupguid=
set remoteopenguid=
set ofile=%temp%\cfop.txt

if exist %ofile% del %ofile%
if exist %ofile% goto ERROR1

powercfg /l | find /v "Existing Power Schemes" | find "*" > %ofile%
for /f "delims= tokens=2" %%i in (%ofile%) do set currentpowerscheme=%%i

echo Changing current power policy (%currentpowerscheme%) to allow automatic sleep with remote opens...

powercfg /l | find "*" > %ofile%
for /f "tokens=4" %%i in (%ofile%) do set currentpowerschemeguid=%%i

powercfg /q | find "(Sleep)" > %ofile%
for /f "tokens=3" %%i in (%ofile%) do set sleepsubgroupguid=%%i

powercfg /qh | find "remote opens" > %ofile%
for /f "tokens=4" %%i in (%ofile%) do set remoteopenguid=%%i

Powercfg -setacvalueindex %currentpowerschemeguid% %sleepsubgroupguid% %remoteopenguid% 1
Powercfg -setdcvalueindex %currentpowerschemeguid% %sleepsubgroupguid% %remoteopenguid% 1

powercfg /qh %currentpowerschemeguid% %sleepsubgroupguid% | more +66 | find "0x00000001" >nul
if errorlevel 1 goto ERROR2
echo Success!
GOTO END

:ERROR1
echo cannot delete %ofile%
TYPE NUL | choice /c:5 /td,10 >NUL
goto END

:ERROR2
echo Cannot change policy (ensure adequate permissions)
goto END

:END
set currentpowerschemeguid=
set sleepsubgroupguid=
set remoteopenguid=
set currentpowerscheme=
set ofile=
REM Pause 3 seconds so user can see messages
CHOICE /T 3 /C ync /CS /D y >nul

:quit

Reducing IT footprint, 15 PCs at a time

Mark Aggar — Fri, 04 Dec 2009 22:39:00 GMT

What’s better than a classroom full of PCs? How about 1 PC per classroom? As backwards as this may sound, it’s exactly the promise of the recently announced Windows MultiPoint Server 2010. The simple goal of this solution is to reduce the cost of providing a rich Windows PC experience by allowing many users to use one (powerful) PC simultaneously. Simply add multiple monitors and USB devices (keyboards, mice, speakers), and with a bit of software magic built upon Terminal Services technology (or Remote Desktop Services as it has recently been renamed), the PC management experience for school teachers everywhere just got a lot more simple.

I had the pleasure of a demo of the solution a couple of months ago, and it is a dream come true. I mean that almost literally – when I first joined Microsoft back in 1999, I was the product manager for Terminal Server. In 2001, the Terminal Server architect showed me a hacked up system that allowed multiple users on the same machine, using similar techniques that MultiPoint Server is using today. The main difference was that in 2001 we were using a Pentium 4 and not very much RAM by today’s standards. As is the fate of many prototypes, this one didn’t survive the prioritization process, but I always held out hope that one day it would see the light of day.

To cut a long story short, and in no small part to some bright folks from Microsoft Research India and support from the Unlimited Potential Group, Windows Multipoint Server was green-lighted for development in Chief Strategist Craig Mundie’s Startup Business Group.

The demo itself was impressive. 16 monitors simultaneously playing 720P HD video using the processing power of one PC (an Intel Core i7 system). This is particular feat was aided by a new feature available in Windows 7 and Windows Server 2008 R2 that allows for full screen video rendering within a remote TS session. If you have two Windows 7 PCs, you can try this for yourself - just open up a WMV file from within a remote desktop session on another W7 PC and prepare to be amazed.

In case you hadn’t guessed by now, this is a win for the environment in several ways. First and foremost it helps reduce waste by dramatically reducing the number of processors, disks and RAM and other electronic components required to serve a given number of users. Secondly, it also significantly reduces the amount of power required to support those users. And third, not only is there less e-waste to deal with at the end of the day, but it makes it far more feasible (financially and practically) to upgrade the central PC’s components (e.g. bigger disks, more RAM) than it would be for multiple PCs – quite often it’s just easier and ultimately cheaper to replace the PCs altogether. The cherry on top is that no special hardware is needed – the whole thing runs using standard (multi-port) video cards and USB hubs, so if you decide you don’t need the system anymore, it’s easy to repurpose everything.

I think MultiPoint Server has the potential to be very successful, especially in school districts where for one reason or another, students don’t have access to laptops. It also makes a lot of sense in Internet Cafe’s and libraries. And it doesn’t hurt that it’s easier on the environment and your energy bill as well.

For more information on Windows MultiPoint Server 2010, check out the Unlimited Potential team blog post.

Better Power Management in the Enterprise

Mark Aggar — Tue, 27 Oct 2009 20:18:00 GMT

Tales of PC Insomnia

Mark Aggar — Wed, 21 Oct 2009 23:20:00 GMT

[note: An update on a couple of the issues mentioned here can be found in my follow-up post here.]

Sleep is a wonderful thing. Particularly if it saves you energy. Of course I’m talking about PC sleep (although I suppose it applies to your own sleep as well). If you’re using Windows XP, chances are your PC isn’t sleeping automatically unless you manually configured it to do so. However if you’re using Windows Vista or Windows 7, you might have noticed that your PC will sleep automatically after 30 minutes or so. If you are running Windows on a work PC, you may have discovered that your IT manager turned off automatic sleep, making you responsible for turning your computer off.

Turning on automatic sleep takes care of this “chore” for you, and it’s easy to configure in the power control panel. Fortunately, many IT managers are revisiting their decision to disable automatic sleep given the potentially large savings that can be had. However, if group policy settings prevent you from configuring your PC to sleep automatically, then you can at least use manual sleep instead of shutting down Windows to save power. On modern PCs, sleep uses only a tiny bit more power than shutting down, but resumes much faster than booting from cold.

Windows 7 PC Insomnia Diagnostic tool

However, there are times when mysteriously your PC won’t go to sleep automatically, even though you can put it to sleep manually. This is phenomena is commonly known as “PC insomnia”. Up until Windows 7, it was pretty much impossible for the average user or even IT administrator to determine why a PC wouldn’t go to sleep automatically. Fortunately Windows 7 includes a new built-in tool to help you diagnose PC insomnia.

In Windows 7, powercfg.exe has been enhanced with a couple of options to help diagnose power management problems, including sleep issues. One of these options is the ‘/requests’ switch which will tell you what components are requesting that the system not go to sleep automatically. The following is output from the powercfg tool using the /requests switch.

C:\Windows\system32>powercfg /requests
DISPLAY:
[PROCESS] \Device\HarddiskVolume2\Program Files\Microsoft Office\Office12\POWERPNT.EXE

SYSTEM:
[PROCESS] \Device\HarddiskVolume2\Program Files\Microsoft Office\Office12\POWERPNT.EXE

AWAYMODE:
None.

(you’ll need to do this from an elevated command prompt – type command at the start menu search box, in the search results, right click on “Command Prompt” and chose ‘Run as Administrator’)

In this particular example, Microsoft Office PowerPoint is running in slide show mode and is asking Windows to not automatically turn off the display or put the system to sleep. This is perfectly reasonable, because having your computer go off in the middle of a presentation is not particularly helpful! Similarly, if you install an update from Windows Update, the updater process will keep the machine awake until the installation is complete to ensure it doesn’t go into sleep in the middle of an update.

Unintended Consequences of not using Power Availability Requests and Presentation Mode.

Conversely, if you’ve ever wondered why your computer display sometimes blanks during the middle of a presentation or while you’re watching a video in a web browser, it’s because the application is not explicitly making a power availability (PA) request to keep the display awake during the activity. Unfortunately, there’s no reliable way for the computer to know that you’re passively watching (or listening) to the computer, so it is usually up the the application to make the appropriate PA request to prevent the display from dimming or the computer from sleeping. Even then, keeping the display/computer on might not make sense for the majority of scenarios - for instance, you really don’t want Excel to to keep the display or system awake on the off chance that you are presenting your latest sales numbers.

One option for long running videos and presentations is to put the PC in presentation mode. To do this, open the Windows Mobility center by holding down the Windows logo key and pressing X (Windows+X) and choose Presentation mode. This runs a small app that will raise a PA request to prevent the display blanking or the system sleeping. Presentation mode is smart enough to turn itself off if the system is manually put to sleep, so the next time the system wakes, it will go to sleep automatically as usual. Just remember to turn off presentation mode when you’re done.

Causes of PC Insomnia

So what about the occasions when there is no obvious reason why your machine stays awake? Sometimes the reason is justified – you just don’t know what it is – other times the app is just badly designed (although typically not deliberately).

A few examples illustrate these two cases.

Opening Files Across a Network

The first is an example of a good reason your PC should stay awake. When a file is opened in read/write mode on a remote file system (e.g. you open a PowerPoint presentation that resides on remote file server share), the local SMB (or CIFS as it’s sometimes known) redirector will raise a PA request to keep your machine awake so that the chances of data loss are minimized – having the machine go to sleep while the file is open could possibly corrupt the file. Similarly the SMB server on the remote machine (assuming it is a Windows machine) will also raise a power availability request until the file is closed.

While this makes perfect sense, in many situations the end user probably isn’t going to be aware that this particular act (i.e. editing a file that resides on a remote machine) is what is responsible for keeping their PC awake.

This can be a leading cause of insomnia for PCs, including those running Windows Vista. Vendors such as 1E and Verdiem have recognized this and provide sleep overrides which will automatically save files and put the computer to sleep, and notify the user of the action taken when the machine is turned back on.

While it may be annoying to some folks that a machine with open files across the network doesn’t sleep automatically, this is in fact a prudent and rationale design decision. You really don’t want your PC potentially “auto-corrupting” files by going to sleep automatically, or people will turn off automatic sleep entirely.

However, this behavior is application dependent. For instance, Microsoft Word 2007 copies the file to the local disk and manages any edits there, only writing the contents back to the original location during a save operation.

In Windows 7, this problem has been addressed somewhat by changes to the way offline files/folders work. Offline files allow users to work with server-based documents even when they are not connected to the server by making a cache of the documents (typically the whole folder) available locally. (Time for a PSA here: If you are an IT administrator in charge of ensuring that PC data gets backed up, Offline files has been enhanced in Windows 7 and is a great solution.) By ensuring that files are always edited using the cached version, Windows 7 can prevent the insomnia caused by users that leave edited server-based documents open.

[Unfortunately there appears to be a issue in the way Offline files treats the temporary files created by PowerPoint and Excel, so opening a PPT or XLS file from a folder marked offline still causes a PA request to get raised. I’m working with the Offline files and Office team and we’ll hopefully see this fixed soon.]

Windows 7 Search Indexer and the Remote File System Redesign

There are other examples where application design can inadvertently keep PCs awake by causing PA requests to get raised for no good reason. One particular example we discovered during the development of Windows 7 involved the Windows search indexer. When adding a remote folder to a local “library” (a very cool feature in W7 that makes it much easier to group and search across your various files regardless of where they are stored), the search indexer opens a named pipe to the remote indexer (assuming the indexer service is running on the remote machine) to keep track of the files on the remote machine. As a result of this action, a PA request was raised – forever. This meant your machine would never go to sleep automatically. It turns out the the Windows 7 file system redirector was raising the PA request in response to the remote named pipe being open.

After discussing this issue at length with the indexer and remote file systems teams, it was determined that it was bad design to have the redirector arbitrarily raise PA requests whenever a remote named pipe was open, simply because there was no way to know whether the application really needed the machine to stay awake. Instead the application that opens the named pipe should raise the PA request if need be. In the case of the indexer, it was determined that it’s not necessary to keep the PC awake, so no action was taken on their part (i.e. fixing the named pipe code meant that the indexer was fixed for free). The good news is that was addressed in early on for W7 RC1, so you can add as many network shares as you like to your W7 libraries and be confident that this won't keep your PC awake.

Adobe Flash Video

Probably one of the most common causes of insomnia on consumer PCs is Flash Video, in use on many web sites (e.g. Youtube, Hulu, MSN Video). Flash has the unfortunate characteristic of streaming silence to the audio stack when the video finishes or even if the audio/video stream is paused. Streaming audio (even ‘silence) through the Windows audio stack causes a PA request to be raised because its assuming you don’t want the PC to go to sleep automatically while it is playing music. There’s not much you can do about this right now so the best advice right now is to make sure you close any web pages that have Flash video on them after you have finished viewing. This problem also manifests itself with some web display ads as well and you don’t need to be explicitly interacting with the ad in order to be affected by this. Unfortunately there is no word from Adobe on when they will address this problem, but I am hopeful that we might see an update from them before the end of the year.

Conclusion

As you can see there are sometimes legitimate and non-legitimate reasons why your PC experiences “insomnia”. Any application (including built-in Windows components) can purposely or inadvertently (if not designed correctly) prevent your PC from sleeping automatically. With Windows 7, Microsoft has added new capabilities to help diagnose these issues, and continues to work to ensure that its own applications are being good “power citizens”.

Third parties also have a significant responsibility here as well. Fortunately, there is plenty of guidance available for developers on microsoft.com to help make systems, applications and devices more power aware, including specific advice on when and how to raise power availability requests.

Part 2 of this topic is here.

E-Readers – better for the environment than paper books?

Mark Aggar — Wed, 30 Sep 2009 03:46:50 GMT

A co-worker recently sent a message to our team asking whether an Amazon.com Kindle would be better for the environment than books. Of course, with anything related to technology (and me), the answer is, it depends.

My short answer was that if you buy a lot of NEW books, then a Kindle is most likely going to have the smaller environmental impact overall. By how much depends on how many books you read and how long you keep the Kindle. The significant factor is likely to be the amount of resources/energy that go into the manufacture and eventual disposal of the device (Amazon does offer to recycle the device), rather than the energy it uses to read the book. E-readers like the Kindle use very little power (practically only uses juice each time the page is ‘turned’ or data is transferred over the network), so in order to know exactly, you’d have to do a full lifecycle analysis of the Kindle and a book.

Of course if you only borrow or buy used books then that’s going to win hands down. Unless of course we end up in a world where the only reason they print books in the first place is for people without e-readers. :-)

Ecolibris has a great set of links that dig deeper into the issue E-books vs Paper Books – it looks like the E-book has the edge still (with the above caveats) but only a full LCA is really going to give us the real answers.

Wal-Mart to the rescue perhaps? :-)

Disposing of Dead Batteries? Not so fast...

Mark Aggar — Tue, 29 Sep 2009 21:54:00 GMT

[It’s been a long while since my last post – it’s good to be back!]

I replaced some alkaline batteries in my son's "activity seat" the other day. As am I allergic to waste, I habitually test the batteries I replace to see if they really are dead. As I've noticed before, in this case only one of the three batteries was completely dead. One of them had a full charge and the other had at least 50% of its useful charge left. As a result, I only needed to replace one battery, although I expect that most consumers would have habitually replaced the whole set. At scale I suspect this represents a lot of good batteries disposed of every day.

I tend not to throw away dead alkalines either. While I've not experimented with alkaline battery chargers, I keep them in a bag and drop them off at local recycling events when I get a chance. Apparently batteries manufactured after 1996 are supposedly safe enough to send to the landfill (aka 'trash'), but to me this seems rather wasteful and a missed opportunity for an enterprising battery manufacturer, as Chris Jordan's images so vividly illustrate (battery image is 2nd from the bottom of the long set of images).

And while I normally use rechargeable batteries for the most frequently and heavily used devices and toys, as you might have guessed, I'm not opposed to using alkaline batteries for devices that don't use much energy and/or go for months or even years between replacements. And for the D cells in the baby seat, I have so few devices that use that type that investing in a special charger hasn't seemed worth it.

Based on the battery strategy I’ve refined over the years, I recommend the following approach to ‘living with batteries’.

1) Buy a battery tester and test batteries before you throw them out. You’ll likely save money very quickly. You get get a cheap one that just tests voltage or a more sophisticated one that actually tests the load and provides more accurate results.

2) Buy a good battery charger. Fast chargers aren’t always kind to your batteries. I use the LaCrosse Technologies charger and have had good results. You can vary the charger voltage depending on how much of a hurry you are in.

3) Invest in rechargeable batteries for frequently used devices (such as digital cameras and favorite toys). In particular, buy rechargeable batteries that don’t go dead if they sit idle for a few weeks. The Sanyo eneloop AA batteries are pre-charged and hold their charge much better than traditional Ni-MH batteries, plus they are environmentally friendly (so they say!).

4) For frequently used devices that need bigger batteries, buy battery adapters that convert AA batteries to fit in devices that need C cell or D cell batteries and use them with your rechargeable batteries.

5) For devices that use batteries very slowly because of low drain or infrequent use, I contend that the most cost effect and environmentally friendly choice is good old fashioned alkalines like the Duracell Procell.

6) Put all of your old batteries in a big “ziplock” bag and take them with you to a local recycling event. At Microsoft, we have onsite battery recycling via a company called Veolia, so I bring mine into the office once in a while and drop them in the designated bin.

Hope that helps!

The Rise of AV Power and Green TV

Mark Aggar — Fri, 20 Feb 2009 00:41:00 GMT

When I first moved to the United States from England some fourteen years ago, my TV and movie viewing experience consisted of a 21" standard definition TV (with cable!), a VCR and, if I could be bothered to turn it on, a stereo amplifier that was connected to the TV.

How times have changed. Today, my wife and I enjoy High Definition (HD) video on a 52" LCD TV, with video sources from a Comcast dual tuner HD DVR, XBOX 360 Media Center Extender and a Blu Ray player. Oh, and don't forget the AV Receiver that is capable of blasting 7.1x140 watt channels in Dolby TruHD.

Sounds pretty awesome, right? Compared to my old setup it certainly is. However, while many of the components are Energy Star-rated, the dark side is that it's consuming quite a bit of power as well. But maybe not as much as you think.

TV Power

Lets start with the TV. The good news is the TV I'm using (a Samsung LN52A850) is an Energy Star-rated LCD model that only uses about 170 watts of power in "eco power" mode. Equivalent sized plasma TVs can consume twice that or more. Fortunately, most new TVs consume virtually no power (less than a watt) when in standby mode, although I've seen some rather egregious cases of TVs consuming more than 70W in standby. CNet's reviews with their handy 'juice box' feature is a useful resource when evaluating TV power. At CES 2009, manufacturers were touting their green credentials and showing off TVs that used even less power, including a unit from Sony that turns itself off if no-one is in the room.

Now when I say that my TV 'only' uses 170W, we need to compare to where TVs used to be. To take just one data point, my 10+ year old 27" CRT TV pulls about 110W when powered on, which might sound better, but it also uses about 20W when in standby. This can really add up. If both TVs are on 5 hours a day and in standby for the rest, then my LCD would use about 850 watt hours, whereas my old TV would use about 930 watt hours! Over a year at 10 cents a kWh, that equates to about $31 and $35 respectively. While the savings are not very compelling in order to make a case to your spouse for new TV, you could save 380 watt hours a day by using a power strip to cut power to the CRT TV when you're not watching it. However, this sometimes isn't very convenient (that's what remote controls are for, right?), it takes a while to get into the habit of turning off the power strip, and many of the late generation CRT TVs will enter their startup mode every time it is turned on after power is removed.

Cable box vs Media Center vs Media Center Extenders

Now onto cable. Back in the old days, you used to just hook the cable directly to the TV which didn't consume any additional power. That all changed with the advent of 'digital' cable, and more recently with digital video recorders (DVRs). The Comcast HD DVR (Motorola 6412) apparently uses over 40W all of the time whether it's on or in 'standby'. This equates to 350 kWh used in one year, about half the average home's monthly power usage (and about $35). A Windows Vista Media Center PC on the other hand will use a little more, say about 75W when it is on, but its standby mode will only consume about 3W. Assuming it's on for 10 hours a day (5 hours watching and 5 hours recording), that's about 290 kWhs per year, 60 kWhs less than the cable box.

If you're using a Media Center Extender such as an XBOX 360, then the math can get a little tricky. The worse case scenario is when your Media Center PC records a program and then you watch it later. This means the PC is on to record the program and then both the PC and the XBOX 360 are on to watch it. The current generation of XBOX uses about 119 watts in media center, so to watch a program that was recorded earlier, this can add up to over 269 watts, and that's not including the TV or the amplifier power. In my setup, watching an episode of Heroes on via the XBOX extender and audio through my AV receiver uses over 700 watts, vs 400w using the Comcast DVR. But you can have a lot more storage for HD content on the Media Center PC, plus you have access to your shows on multiple TVs throughout the house (with additional extenders) and can make the shows portable (very handy for when you are on vacation with young children).

Wake-on-LAN

The silver lining on using an XBOX 360 as a media center extender is that the 360 can wake your sleeping PC when you want to watch media via the extender functionality. You can do this by enabling wake-on-LAN "magic packet" functionality on the network card of the PC. Instructions on how to do this are available here. When the media center extender software is started on the XBOX, the XBOX 360 automatically sends a wake-on-LAN "magic packet" to wake the sleeping PC. This way your PC only has to be on when you are using it (locally or remotely) or when it is recording a program. Even if the XBOX displays a message that it cannot connect, when the PC is ready the XBOX will connect without any further user intervention. It's fairly family-proof - my wife uses it all the time.

AV Receivers

So what about the AV receiver? I just replaced my 7 year old Sony receiver to a new Onkyo TX-SR876 (also Energy Star rated), principally so I could utilize HDMI for my components (the old receiver is now downstairs). The receiver is very efficient when in standby (less than 1W) and doesn't come on automatically after a power outage like my old Sony receiver does!

The rub with AV receivers is that if you connect your components via HDMI to the receiver (and connect the TV to the receiver), you are going to be forced to watch your components with the receiver turned on. Given that the receiver and subwoofer use a fair bit of energy while on, I wanted to have the option of saving energy while watching TV shows that didn't justify the added energy cost of surround sound. As a compromise I have connected the cable box and the XBOX 360 directly to the TV (utilizing all three of the rear HDMI inputs), and connected digital audio (coax or optical) from these components to the receiver. Only the Blu-ray player is connected directly to the receiver. While this works pretty well with an advanced remote control (see below), this isn't particularly easy and kind of undermines one of the main value propositions of HDMI - i.e. only needing one cable.

Green TV

To enable ease of use, I have a different 'activities' programmed on my Harmony Remote to watch TV and use the XBOX - three with the receiver on and three "green" activities that utilize the only the TV Speakers (and leave the receiver off). I didn't have this option with my old TV and receiver, so ironically this new equipment allows me to use it less and save energy. (For the record, while I generally like my remote, next time I would choose something like the Harmony 890 for its RF capabilities - getting "line of sight" to components can be difficult, and frankly the touchscreen on the Harmony One isn't ideal - I think hard buttons are probably better).

As it happens, both my TV and AV receiver support a new standard called HDMI-CEC. This feature allows you to control the receiver via the TV, so you could just use the TV remote if you wished. I noticed a short disclaimer in the manual that the use of this feature would significantly increase standby power on the receiver. This peaked my interest, and it turns out that they are right - with the feature enabled, it uses about 80W of power on standby, or 700 kWh's a year, about a month's worth of electricity usage. Kudos to Onkyo for pointing that out in the manual, but suffice it to say, that feature will not be enabled in my house!

Also, the Sony Blu-ray player has an option to start-up faster, but at the cost of using more energy on standby (again the Blu-ray player tells you that it is going to use more energy). I haven't measured the energy use of this fast startup mode, but given my previous measurements, it's not something I'm going to be enabling in a hurry.

In Conclusion

So what's the upshot of all of this?

Needless to say this is a mixed bag. It's been clear for quite a long time that our penchant for higher fidelity and personalized entertainment is going to have ramifications for our energy bills. However, if configured optimally these newer devices can use little to no energy when turned off and may actually in aggregate consume less power than older devices. Cable and Satellite companies are hopefully taking note and planning set top boxes that consume near-zero watts when on standby.

Also, while I applaud device makers that are reducing both operational and standby power, there are some worrying signs that these efforts could be quickly undermined by brand new 'user friendly' features such as the aforementioned HDMI-CEC that wipe out any savings. Also, the industry could do more - in particular, "green" AV Receivers that provided a passive HDMI pass-through mode which consumes little-to-no power when active or in standby. This would make it much easier to install and configure systems that allowed families to enjoy hi-fidelity audio when it made sense, vs having the receiver on every time the TV was on, saving lots of energy as a result.

So, how green is your AV experience?

One shiny new gadget, hold the lacerations please!

Mark Aggar — Mon, 29 Dec 2008 01:53:00 GMT

Like most people, I've never been a fan of the so-called "blister-packs" that house today's electronic gadgets. While I've never personally done any serious damage to myself when trying to open one of these, it's understandable how folks can even end up in hospital as they anxiously attempt to free their new "e-joy" from its plastic prison. Added to this aggravation, you typically end up with a bunch of packaging that has no other destination than the landfill. That's why I was very pleased with my experience in opening a new Microsoft Explorer Mini Mouse (with "BlueTrack^TM technology" no less!).

The fact that I managed to open the packaging with no sharp instruments and a 4 month-old baby boy sitting on my lap is no less short of amazing. I'm pretty sure that Liam didn't appreciate the breakthrough in packaging technology as much as I did, but I'm sure he was thankful that I didn't have to put him down to achieve this feat that previously would have been considered negligent parenting. The magic breakthrough is a couple of tabs on the back of the package which you tear down the back, allowing you to open the sides. A good firm tug on the sides then snaps them off and voila, you have free access to the contents.

And the sweet surprise inside? (other than the mouse itself which works on my granite countertops). A note printed on the paperboard which states that the plastic portion if the package is manufactured using PET, not PVC, which means it is #1 recyclable! (the same material used in water bottles). [update: Not so fast: According to our packaging folks, the PET cannot be recycled because it doesn't look like a bottle. Yep, most recycling facilities manually sort plastics and discard anything that doesn't look like a bottle, regardless of the material it is made of. Bummer). Along with the paper and cardboard inserts (manufactured using 50-75% recycled paper and using soy inks to boot), this means that for the first time in my experience, all the packaging can go straight into the recycle bin [well not quite as I explained above]. And the Mouse product managers were even kind enough to put a link to our http://microsoft.com/environment site!

Also, according to this story on NPR's Marketplace, Amazon.com is teaming up with some manufacturers, including Microsoft, to provide "frustration-free" packaging. The story cites a toy that literally just comes packaged in a carboard box with just plain brown corrugate filler that is 100 percent recyclable. Of course this works for Amazon because they don't have to deal with theft by consumers, so what Microsoft is doing with products like computer mice is hopefully a trend that others will follow for the brick-and-mortar channel.

If I had one criticism it would be that there is no indication on the outside of the package or the web site that the packaging is indeed recyclable. If there's anything that consistently gives me buyer's remorse, it's when I end up with a bunch of packaging that I have no other choice but to put in the garbage. As more manufacturers start to adopt more sustainable packaging strategies (and really, I don't want to see another polystyrene foam insert or pellet as long as I live), I think this will become a differentiator for their products.

[so I guess I can still put the PET parts in the recycling and hope they make it past the sorters]

What do you think? Does or will packaging affect your buying decisions?

Electric Cars = Environmental Paradox?

Mark Aggar — Tue, 16 Dec 2008 02:58:00 GMT

I was just reading a BBC news story about the new electric Mini E. This little beauty does 150 miles on a 2 hour charge at 48 amps (presumably on 230V). A quick back of the envelope calculation (actually calc on my Windows Mobile smartphone) yields figures of about 7 miles per kWh, or at 10 cents a kWh, about just 1.5 cents a mile. This is admittedly a lot better than the quoted “half the cost of a gas powered Mini”, so BMW may be considering charging you less for the vehicle (e.g. subsidizing the expensive battery) and making up the difference by effectively charging you more for electricity, i.e. miles driven.

Either way, does anyone think that a vehicle that costs half as much or less to drive is going to sit in someone’s garage while they take the bus? Especially a vehicle that isn’t exactly cheap to buy in the first place (subsidies or not, the BMW Mini field trial is $850 per month for 1 year), and is just begging to be driven - both because it’s fun as heck and also because if you’re going to be buying all that embodied energy, the only way of bringing down it’s true cost per mile is to drive it – a lot.

Which brings me to the paradox. Electric cars, while arguably better for the environment, are poised to make traffic congestion worse for everyone because they are cheaper to operate and thus will be driven more. Consequently, the remaining fossil-fuel powered vehicles on the road are going to be stuck in traffic more often and produce more carbon as a result.

Think about it - the owners of these vehicles will feel that they are doing the environment a favor by driving an electric vehicle, especially if it's powered by renewable energy. And frankly, if they don't drive it a lot, it seems that a lot of energy and materials would be locked in their garage for no good reason.

So how do we solve this conundrum? As you might expect - software is the key ingredient. Alongside the oft-mentioned smart-grid technologies that would allow the vehicle to be charged when electricity generation is plentiful and green (e.g. windy nights and sunny mid-mornings), additional software and systems can be used to reduce traffic congestion in very efficient ways. Yes folks, I’m talking about car-pooling - Twenty-first century car-pooling software to be exact. Specifically, software that allows drivers to find - in real-time - other people who are going to destinations along their route – effectively reducing the amount of vehicles that need to be on the road at any one time.

The challenge with traditional car-pooling (or ridesharing as it’s known in the transit profession) is that it requires people to pre-arrange their travel, which invariably means that you are going to have to leave and return at the same time and place pretty much every day of the week. While this might work for folks who work on the clock, it isn’t particularly useful for large segments of the population that have some variability or flexibility in their travel schedule. And it hardly ever works on non-work days, unless you’re including taking the neighborhood kids to soccer practice.

As you travel to work during the rush hours, take a quick look around. Notice all of those empty seats in the cars around you? That’s inefficiency at work. What we need are dynamic systems that can fill those empty seats and reduce the number of shiny metal boxes traveling to any one location at any given time.

A smart, dynamic car pooling system can fix this. Leveraging the power of social networks such as Facebook and Myspace, or even the fact that you are likely to (initially) trust someone who simply works in the same company as you, ridesharing matches can be made automatically based on where you are going and when. Many of us are extensively using calendaring and scheduling features offered by applications such as Outlook and Windows Live Hotmail, and it’s conceivable that this information could be extended and leveraged to provide hints to the system about where you are and where you need to go next.

The key is – the less you have to do to make a ridesharing match, the more likely you are to make one in the first place. Let me use a personal example. My wife and I used to work in locations about 400 yards from each other. But in 4 years, we car-pooled together probably less than a dozen times. Why? Because we both had flexible schedules, and the overhead of trying to figure out if our schedules matched on the way out and back (which they rarely did, especially after we had kids) was just too much to deal with on a daily basis. Is this pathetic? Probably, and I’m not proud. But it’s also reality for most people.

However, by integrating users' calendars into the system, riders can be automatically notified of opportunities to rideshare and accept or decline on the spot using their mobile phone. Similarly, drivers in smart (electric or not) vehicles can be notified via their on-board navigation system that there is someone to pick up along their route, even after they have left home. Incentives for drivers to carry additional passengers can be created, with a debit and credit system for passengers and drivers respectively (potentially as an alternative to road tolling). GPS technology can guide both the passenger and driver to a pick up spot which doesn’t require the driver to go too far out of their way, but is not too far for the passenger to walk. I’m just scratching the surface here.

The best part about the whole system is that you can catch a ride with one person when you leave and a ride with a different person when you return, regardless of the time you travel. Many employers offer a guaranteed ride home program in the event of someone being stranded, which can help provide the safety net such a system needs until a large enough user base is established.

Most important is the need for any given metropolitan areas to invest in a single system (using open standards) so a critical mass of users can be built quickly. In the long run it does no-one any favors to have many competing systems that fragment the user base, thus reducing the chance you will find a match and discouraging users from using the system.

So while electric cars are much more efficient at converting energy into VMT (Vehicle Miles Travelled), we need to look at the personal transportation challenge holistically. Simply replacing most gasoline-powered passenger vehicles with electric ones is probably necessary, but definitely not sufficient. We need to reduce the number of cars on the road as well. A system such as this could achieve that and also reduce the number of vehicles each family owns, because you’d only really need a dedicated vehicle if all the family was going out together or if you had a large amount of cargo to transport. Taking that further, if you integrated the system into car sharing schemes (like ZipCar), many families who own multiple vehicles today might be able to get out of vehicle ownership altogether.

The reality is we just don’t have enough space to keep on building roads to deal with an increasing number of single-occupant vehicles. Jevon’s paradox proposes that as you make something more efficient, the rate of consumption increases. Cheaper-to-run cars are very likely to mean more cars on the road - unless we fill most of the seats in the process. I doubt that more traffic jams would be considered great progress, even if we aren’t spewing as many poisonous gases and CO2 into the atmosphere while we sit there. The silver lining is that electric cars are very efficient when not moving or moving very slowly. However, most of the liquid fuel-burning vehicles likely to be all around you are not.

So, when you start looking for your next new vehicle (assuming you can obtain a line of credit!), ask the dealer what sort of systems come with the car to help you fill all the seats on each journey. Sure, you are likely to get a blank stare in return, but miles per gallon (or MPkWh - miles per kilowatt-hour) shouldn’t be the only measure of efficiency if we are to have a road-based transportation system in the not-too-distant future that isn’t gridlocked for most of the day.

And besides, when you’re in your brand-spanking new shiny electric car, aren’t you just dying to show it off to someone else? :-)

P.S. In case you were wondering, the picture depicts a Model T Ford and a Chevy Volt electric battery pack.

Server Efficiency Pot o' Gold

Mark Aggar — Thu, 11 Dec 2008 02:11:00 GMT

My colleague Matt Robben from the Windows Server performance team has produced a very comprehensive white paper and accompanying blog post which goes into great detail about how you can save energy with Windows Server 2008 power management, including new options for increasing savings even further (up to 10% more than the default configuration, based on the systems the performance team tested). Even if you aren't running Windows Server 2008, the white paper is chock full of useful advice on how to save energy through your hardware configuration and the components you use. If you operate, spec or buy servers, this is a must read.

The paper correctly points out that the cost to power and cool the server is the same if not more than the cost of buying it when amortized over three years. With this in mind, you’re probably thinking that any investment in power efficient hardware is worth the potentially extra cost. Generally it is, but there are some exceptions.

For instance, based on some analysis I did a few months back of SPEC power benchmarks, it turns out that buying low power processors might not make much financial sense - depending on the price you pay. I calculated that with a $200 spread between two equally performant processors with different TDPs (Thermal Design Power), the ROI is about 7 years – obviously way beyond the life of the machine. However, the $200 difference in question was based on 'list' prices. On Amazon.com (they really do sell everything!), the difference was only $30 – which had a payback of about 7 months for a single processor.

This obviously applies to other components as well (e.g. price per GB for 2.5” vs 3.5” disk drives). If more efficient components are more expensive, then it might be worth doing some rudimentary math to see if the power savings will ever exceed the increased cost of the components before the machine is retired.

Efficiency part 2

Mark Aggar — Wed, 10 Dec 2008 04:04:00 GMT

A few things today.

First, part 2 of my Green IT interview has been posted on TechNet Radio. Direct links are available to WMA, MP3-Hi and MP3-low. In this one, I talk predominately about virtualization.

Also in the deja vu category is this TechNet Edge video interview on our new Gen 4 Data Center architecture with David Gauthier, Data Center Infrastructure Architect and Christian Belady, Principal Power and Cooling Architect, from the Global Foundation Services team at Microsoft. They offer a behind-the-scenes architectural perspective on the concept behind modularity and our data center and infrastructure strategies at large. The new blog post discussing the infrastructural components and efficiencies gained from our Gen 4 data center plan can be found here.

Finally, I was forwarded this very interesting interview with Amory Lovins of the Rocky Mountain Institute about efficiency being our best untapped energy resource. Amory is the guy who typically carries around his carbon fiber 'hat' to demonstrate the strength of the material as a substitute for steel in the auto industry. I love this quote from him at the end: "Then I took a lot of video conferencing, rather than flying. Just move the electrons and leave the heavy nuclei at home."

Green IT TechNet Radio Interview

Mark Aggar — Tue, 02 Dec 2008 23:06:00 GMT

My former colleague, Adam Bogobowicz interviewed me recently on Green IT for TechNet Radio. This is part 1 of a 2 part interview. You can access part 1 of the interview via the following links:

WMA MP3 (Hi bandwidth) MP3 (Low Bandwidth)

(Tip: if you are listening via Windows Media Player, hit CTRL-SHIFT-G and I'll talk at 1.4 times the speed to save you both time and maybe some energy. :-) CTRL-SHIFT-N reverts back to normal speed)

During the interview I talk about PUE, EPEAT, Desktop Power Management and Data Center Best Practices among other things. Coincidentally, our Data Center Operations team announced their vision for next gen data centers, which you can read about on Mike Manos' blog. Look ma, no roof!

Greening your PC

Mark Aggar — Thu, 13 Nov 2008 04:54:00 GMT

I’m often asked what can be done to reduce technology’s impact on the environment. Before I answer that, I usually like to point out that technology can have a net-positive effect by reducing the need for transportation and materials and generally making things more efficient. However, there is a lot of opportunity to reduce the amount of energy that technology consumes while it is operating, and to also support more environmentally-friendly manufacturing and disposal practices. For instance, according to the Natural Resources Defense Council, PC power management such as those built into Vista operating system can save you about $50 each year for every desktop computer. Used widely, power management tools could shave $500 million off the nation’s energy bill and eliminate 3 million tons of global warming pollution.

Here are some tips I use when considering my own purchasing and use of PCs and monitors, and I hope you find them helpful.

1. Consider the PC and Monitor you are purchasing:

Laptops vs Desktops – If you’re concerned about saving energy, choosing a laptop over a desktop PC might be your best bet, even if you plan to use it with an external monitor and keyboard. Many modern laptops consume less than 30 watts when running at full performance, compared with a modern desktop PC that idles around 60W and can consume in excess of 150W at full tilt -and that’s not including the monitor. Many new small form-factor laptops idle at less than 15W – less than the power used by a typical Compact Fluorescent Light (CFL) bulb.

ENERGY STAR rated PCs and Monitors – These days it’s almost impossible to find a monitor or PC from a reputable manufacturer that doesn’t have an ENERGY STAR label on it. ENERGY STAR rated devices have earned their rating through meeting a rigorous set of guidelines and qualified products must now meet energy use guidelines in three distinct operating modes: standby (off mode), sleep mode, and while computers are being used. Also, qualified computers must also include a more efficient power supply (typically an 80 Plus rated version). While ENERGY STAR is a United States government standard, countries around the world are adopting the standard. Most people don’t realize though that the ENERGY STAR requirements are getting tougher, so the energy savings on newer PCs are likely to be better than a similar model from previous years.

LCDs vs. CRTs – An LCD monitor can consume half to two-thirds as much power as the an equivalent-sized CRT monitor, which is great news if you are in the market for a new display. However, many folks tend to ‘trade-up’ when the buy a new monitor, so potentially you may be consuming even more energy than you were with a small CRT. Along with checking how many watts the monitor consumes when it is active, simply turning down the brightness can save a significant percentage of the monitor’s energy use also. Plus your eyes will thank you after a long day of staring at the screen!

Look for EPEAT – The EPEAT (Electronic Product Environmental Assessment Tool) database evaluates electronic products (desktops, laptops and monitors) on 51 total environmental criteria and can help identify environmentally preferable electronic products. The three-tiered EPEAT rating system includes 23 required criteria and 28 optional criteria. The optional criteria are used to determine if the equipment receives EPEAT Bronze, Silver, or Gold recognition. In addition, due to EPEAT’s requirement that registered products meet ENERGY STAR specifications, these products will consume less energy throughout their useful life. EPEAT was originally developed for institutional or business line purchasers, so many PCs targeted at individual consumers are not yet EPEAT rated, and recycling services provided by the manufacturer for EPEAT-rated PCs and monitors may not be available to individuals. EPEAT is a US-based rating but has been adopted by countries such as Canada, Brazil, New Zealand and the United Kingdom.

2. Consider how you use your PC and Monitor:

Turn on Power Management. Using the power management features on your PC is one of the easiest and cheapest things you can do to save energy. Power management features in the Windows Vista are on by default and will automatically put the PC to sleep after 60 minutes of idle time, and the monitor will turn off automatically after 20 minutes. You can always change the settings to be more aggressive and save even more. For example, my desktop PC at home (which is primarily used as a Media Center PC which I remotely access through an XBOX 360) is set to go to sleep after 10 minutes, and the monitor turns off after 5 minutes.
There are also free tools available to consumers, such as Edison by Verdiem, that help users track how much energy their PC is using and show how much could be saved by enabling various power management settings. In the Enterprise, vendors like Verdiem and 1E also provide comprehensive power management solutions to address your particular organizational needs.

Disable Screen Savers. It should come as no surprise that screen savers don’t save energy, and in some cases use more energy than when you are using the PC yourself. What may be less obvious is that using a screen saver instead of powering down the display may shorten the life of an LCD monitor due to the fluorescent tube becoming progressively dimmer as it is used. So say goodbye to those flying toasters and use display blanking instead (accessible via the power management control panel interface). If you use the screen saver to automatically lock your PC if it is left unattended for a while, I recommend using the blank screen saver (which doesn’t consume additional energy), and setting the monitor blanking timeout to be slightly shorter than the screen saver timeout. This way, if your monitor blanks while you are sitting at the machine, you can move your mouse to bring the display back up and not have to enter your password each time.

Unplug unused equipment when not in use (monitor, PC, wireless phones, printer, etc.). Many of your older electronic products that are plugged into the wall can still draw significant amounts of energy when they aren’t even being used. If an external power supply is warm, then it’s consuming at least some energy. However, most new devices are actually very energy efficient and will only draw a couple of watts or less while they are on standby, including PCs and monitors. Unless you have a watt meter and are able to measure individual devices, having them on a power strip and turning it off when not in use might make a noticeable difference on your energy bill, particularly if your devices are more than a couple of years old. You can buy smart power strips that automatically cut power to your other devices when the PC goes into sleep mode. The power strip provides the few watts the PC needs while in sleep, but can cut power to everything else until the PC is turned on.

3. Conscientiously dispose of your old or unwanted PC and Monitors

Recycle or Donate old PCs and Monitors: There are a number of great ways to handle the disposal of old PCs and monitors. While recycling through a reputable third party PC recycling vendor or directly through a PC/Monitor manufacturer is one good option, another option is to donate them to a technology refurbisher. Refurbishment reduces computer waste and provides affordable technology to millions of people. Today, about 28 million computers are refurbished and resold or freely distributed. Unfortunately, while 80 percent of the world's population is still without access to computers and the Internet, tens of millions of still-usable computers are discarded each year by businesses, individuals and organizations, mostly in the developed world. The reuse of computers offers tremendous promise for increasing access to information, especially for people underserved by technology. In addition, this helps ensure that older PCs don’t end up in landfills and are put to good use. To donate your old PC and to find a Microsoft Authorized Refurbisher near you, visit: http://www.techsoup.org/mar/marList.aspx

Well that’s a lot of advice and I hope you’re not too overwhelmed by it. Becoming energy efficient and environmentally sustainable is a journey and doesn’t have to be an all or nothing proposition. Start with the low hanging fruit first (e.g. enabling power management and buying Energy Star and EPEAT rated devices) and you can worry about fine tuning later on down the road. In a future post I’ll be writing about some of the experiences I have had with the energy consumed by other consumer electronics, such as TVs and Game Consoles.

Welcome to my blog!

Mark Aggar — Thu, 13 Nov 2008 04:48:00 GMT

Greetings!

My name is Mark Aggar and I'm Microsoft's Environmental Technologist. I work on the Environmental Sustainability team, reporting to our Chief Environmental Strategist, Rob Bernard.

Mine is one of the more unique job titles at Microsoft, and my role is to drive thinking (and action!) about how we reduce the footprint of Information Technology as well as the role that technology can play helping to reduce the footprint of other industries and activities (software for sustainability, or Green by ICT). And because Microsoft is such a heavy user of technology I tend to get involved with our own footprint reduction efforts as well.

I'll be using this blog to share my thoughts on technology's role in helping (or hurting) the environment and I hope to have a dialog with you on the various topics I write about.

Suffice it to say that I think I have one of the most interesting and unique jobs at Microsoft and I feel incredibly fortunate to be in a position to make an impact at scale.

You can also check out the official Microsoft Environmental Sustainability team blog at http://blogs.msdn.com/see, where I will be cross-posting occasionally.

-Mark